---
dataset_info:
  features:
  - name: source_id
    dtype: int64
  - name: question
    dtype: string
  - name: response
    dtype: string
  - name: metadata
    struct:
    - name: score
      dtype: int64
    - name: source
      sequence: string
  splits:
  - name: train
    num_bytes: 2380772709
    num_examples: 1143608
  download_size: 1150372955
  dataset_size: 2380772709
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---